A deep learning-based cytological image rapid and accurate identification method and system
By employing deep learning methods for multi-scale feature extraction and dynamic edge enhancement, the problem of misdiagnosis and missed diagnosis in FNAC image recognition has been solved, achieving highly accurate and consistent cytological image recognition applicable to various medical scenarios.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- PEKING UNIV SCHOOL OF STOMATOLOGY
- Filing Date
- 2026-04-22
- Publication Date
- 2026-07-17
AI Technical Summary
Existing deep learning models in FNAC image recognition suffer from insufficient receptive fields and low tolerance to low cell counts and complex backgrounds, resulting in high rates of misdiagnosis and missed diagnosis. Furthermore, they lack diagnostic consistency and information utilization, making it difficult to meet the needs of precise stratification and standardized management.
By employing a process of multi-scale feature extraction, fine-grained optimization of shallow features, multi-scale feature fusion, and dynamic edge enhancement, and by constructing a multi-scale representation fusion network model and an edge dynamic enhancement algorithm, accurate identification of FNAC cytology images is achieved.
It improves the accuracy of benign and malignant identification, reduces the false negative and false positive rates, enhances diagnostic consistency and efficiency, supports rapid pre-screening of batch samples, and adapts to the application needs of different medical institutions.
Smart Images
Figure CN122416441A_ABST